Electronic ignition

Post your post 1970 technical queries here

Moderator: Moderators

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Tue Jan 08, 2019 6:15 pm

So much careful measuring later it still makes no sense to me.

This is the test set up. I am feeding a sine wave into the coil on top of the Boyer sensor:
20190108_145538.jpg
Measuring the time between trigger of the scope and fire from the Boyer so this is relative only, not absolute and it is a time, not an angle. I would need to do some calculations to make it angle absolute and right now I CBA. It is not as I expected.
20190108_145618.jpg
boyer_advance.png
Interestingly it stops sparking completely at 408 RPM. That is dependant on input sine wave size obviously and may in reality work at lower revs than that.
You do not have the required permissions to view the files attached to this post.

vmx1200
Forum User
Posts: 129
Joined: Mon Mar 26, 2018 12:02 pm
Country of Residence: UK
Location: Guildford

Re: Electronic ignition

Postby vmx1200 » Tue Jan 08, 2019 8:49 pm

for what its worth my 78 RS still has points and a centrifugal advance, it starts well as long as the battery is fully charged and accelerates as well as I need and is capable of speeds I certainly don't need.

Not saying it isn't fun to rise to the challenge of developing one's own tech
1978 R100RS owned since 1981
2017 R Nine T Pure

Roy Gavin
Forum User
Posts: 410
Joined: Sun Jan 21, 2018 9:29 am
Country of Residence: Australia

Re: Electronic ignition

Postby Roy Gavin » Tue Jan 08, 2019 11:11 pm

Yes, I dont believe the Boyer curve on the net either, but then I don't believe much of what I read on the net anyway, I am just programmed that way!
I doubt if Airheads have any special needs in respect to ignition curves and a simple analog curve which replicates the action of centrifugal bob weights would probably give results little different to any other.
Any conversations I have had with guys who operate dynos for a living would seem to confirm this, and if you look carefully at pictures of production race bikes they often had manual advance and retard levers long after road bikes had something supposedly more sophisticated!

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Wed Jan 09, 2019 6:41 pm

The rotating magnet has two North poles on the outside, that seems a stupid way to do it to me but it is certainly like that. That means two sine waves per camshaft revolution, one per crank revolution.

This is the advance curve. The starting point is undefined as I have no timing marks on my sine wave but the curve is correct, it may just be out vertically.

It is a stupid curve and I have checked it several times, one reading per 60 RPM (1 Hz increment).
advance_angle.png
You do not have the required permissions to view the files attached to this post.

User avatar
Rob Frankhamr
Club Member 13
Posts: 2973
Joined: Mon Jul 25, 2005 6:33 pm
Country of Residence: Scotland
Location: Kinloch Rannoch, Perthshire

Re: Electronic ignition

Postby Rob Frankhamr » Wed Jan 09, 2019 8:37 pm

That certainly looks like a pretty ridiculous curve... can't see why anyone would make a unit like that.

Have you taken into account that the amplitude of the induced voltage in the coils induced by a moving magnetic pole will change dependant on the rotational speed of the engine? If, as I assume, the trigger point is determined by a comparator measuring the induced voltage against a reference voltage, then the amplitude of the wave will have to be be taken into account because it will affect the timing. Back in the day, the Honda 250 dream series used an ignition system which had two sensor coils of different design producing different amplitude and width waveforms at the same time. These waveforms were compared and when the voltages 'crossed' the ignition fired. The actual ignition advance was determined solely by the comparison of the two waveforms.

The two magnetic North poles used by BB combined with two sensors was, if I remember correctly, to produce an averaging effect to compensate for any slight variation in the position and/or sensitivity of the sensors and/or strngth of the two magnets. The effect is to produce the same firing point for both engines. The two sensor coils are, I think, linked in series.

Rob
Robin Frankham
ImageImageImage

Frankhams retirement home for elderly Boxers.

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Wed Jan 09, 2019 9:50 pm

But with one magnet and the other coil the other way it would be the same. The problem with having two magnets is that they are opposed in the centre where they are stuck to a lump of steel. This will make the magnets last for a lot shorter time than doing it the other way.

I can't compensate for accuracy of the Boyer trigger circuit. I have done some partial measurements with my lathe and it comes out pretty close but I only tested at a few different speeds. I have also run it from a square wave but that doesn't really give any more info.

For sure if I had doubts about the Boyer curve on t'interweb I have even bigger doubts about mine but can't see where I did anything wrong.

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Wed Jan 09, 2019 10:11 pm

Increasing amplitude increases dwell only, no change in trigger point which sort of says it uses zero cross detection which I would expect anyway. Reducing amplitude leads to the unit switching to fixed pulse width rather than having a dwell of 50% as it does when you slow it right down (<500 RPM). Damned confusing. :)

User avatar
Rob Frankhamr
Club Member 13
Posts: 2973
Joined: Mon Jul 25, 2005 6:33 pm
Country of Residence: Scotland
Location: Kinloch Rannoch, Perthshire

Re: Electronic ignition

Postby Rob Frankhamr » Wed Jan 09, 2019 11:06 pm

Maybe... I can't offer any other reasonable explanation. All I can think of is that, with thousands of people using that ignition unit on airheads and on other machines, someone before now would have noticed if the advance curve was a sawtooth... after all it should be obvious using a timing strobe and a Mk 1 eyeball.

Rob
Robin Frankham
ImageImageImage

Frankhams retirement home for elderly Boxers.

Roy Gavin
Forum User
Posts: 410
Joined: Sun Jan 21, 2018 9:29 am
Country of Residence: Australia

Re: Electronic ignition

Postby Roy Gavin » Thu Jan 10, 2019 5:42 am

There were two magnets on the plate in the analog BB unit I fitted to my BSA B44, and they both fired the plug, one wasted.
Unfortunately, one magnet had little magnetism and hence little spark, the other would fire the motor on the exhaust stroke.
Turning the plate 180 degrees fixed the problem and it has worked fine ever since.
So on this unit I doubt if there is any connection - but I no nothing of the working of the current Microdigital unit, except that it has been fit and forget and works flawlessly.
My quick check of magnetism is now on the BB site - the magnet should hold up the 2"x1/4 bolt that holds the plate on, when hung vertically.

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Thu Jan 10, 2019 5:42 pm

Found the problem. I took the measurements off load. I didn't want the sort of RF noise one gets from a working coil. I have done the measurements again with a bulb for the load and it makes a lot more sense.

I am unsure why this should be so and I am not about to de-pot the Boyer to look at the PCB. Interestingly the thing carries on advancing right up to 20,000 RPM, I stopped at that point, presumably it goes on until it is too fast for the electronics.
timing_load.png
You do not have the required permissions to view the files attached to this post.

Roy Gavin
Forum User
Posts: 410
Joined: Sun Jan 21, 2018 9:29 am
Country of Residence: Australia

Re: Electronic ignition

Postby Roy Gavin » Fri Jan 11, 2019 1:09 am

If the revs were halved that it would make a little more sense!

Does it always automatically advance around 8 degrees? or was that your static setting?

But looks pretty much as I forecast in a previous post, near enough replicates bob weights until they weights are all in.

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Fri Jan 11, 2019 7:40 am

I concur.

I keep thinking about whether or not I have the revs doubled but I don't think I have. Same applies to the actual angle, do I have it doubled or halved.

Thinking about this is one of the reasons I am doing the whole thing, it makes my brain work, my brain does not work as well as it did before I had a heart attack.

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Fri Jan 11, 2019 4:00 pm

I am struggling to be sure I have this right. The graph appears to be double revs so can someone with a more functional brain check.

One sine wave in = 1 rev of engine crank (half cam)
16 Hz in = 960 RPM (crank)
At this revs the coil wire goes high 11.6 ms after my scope triggers (arbitrary trigger point, level set to zero)(no timing marks on sine wave)
One crank rev = 62.5 ms (1000 / 16)
Angle to fire at this rev is 66.8 degrees (360 * 11.6 / 62.5)

Subtract a constant 57 or so from every reading to move the graph vertically to start at 9 degrees (static timing).

So, is that correct or do I need to halve or double something?

Roy Gavin
Forum User
Posts: 410
Joined: Sun Jan 21, 2018 9:29 am
Country of Residence: Australia

Re: Electronic ignition

Postby Roy Gavin » Sat Jan 12, 2019 1:54 pm

You have lost me, I just look in the timing hole and make certain the mark changes when the motor revs increase.
Then, if I am in a hurry when I am riding the bike, keep the revs 500 either side of peak torque, which is usually after the advance is all in.
No point on a road bike in revving it any higher, as peak torque is where the motor makes most power, and you will get the best throttle response.

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Sat Jan 12, 2019 3:22 pm

You have lost me, I just look in the timing hole and make certain the mark changes when the motor revs increase.
Then, if I am in a hurry when I am riding the bike, keep the revs 500 either side of peak torque, which is usually after the advance is all in.
No point on a road bike in revving it any higher, as peak torque is where the motor makes most power, and you will get the best throttle response.
I can understand that, I lost myself too. :)

chasbmw
Forum User
Posts: 1239
Joined: Mon Aug 14, 2006 9:19 pm
Location: Bath UK

Re: Electronic ignition

Postby chasbmw » Tue Jan 15, 2019 1:57 pm

I have used Boyers on various bikes since I fitted one on a Commando back in 1973. despite many miles and various climates I haven't had a failure yet.

The reason why a Boyer retards the ignition at around 2500-3000 RPM is because this is the prime Pinging zone for airheads, so retarding the ignition in the zone is a good thing (especially for R60's)

Any Jitter at ultra low revs is the "idle stabilisation" feature, the idea being is to use timing to catch the bike from stalling.

When setting the timing on a Boyer forget BMW instructions and time the bike on the fully advanced timing mark, WHEN THE BOYER HAS STOPPED ADVANCING, typically this will be between 3500 RPM and 4000, but may be higher, so don't look at the rev counter, look at the image produced by your strobe.! If you try to time it at idle you may get confused because of the idle stabilisation (see above)

The current range of Boyers seem to be pretty reliable and have been revised so that they will work well at lower voltages than the previous models, like most electronic ignitions they prefer suppressed plug caps (especially the microdigitals) and benefit from a good earth, proper wiring and a ignition switch that makes good contact.

This is a print out of the Boyer twin plug ignition map that i got from Boyer about 15 years ago.

ImageCharles
Charles
2016 CCM 450
1975 R90/6
1958 R50
2012 R1200R
Image

chasbmw
Forum User
Posts: 1239
Joined: Mon Aug 14, 2006 9:19 pm
Location: Bath UK

Re: Electronic ignition

Postby chasbmw » Tue Jan 15, 2019 1:59 pm

I concur.

I keep thinking about whether or not I have the revs doubled but I don't think I have. Same applies to the actual angle, do I have it doubled or halved.

Thinking about this is one of the reasons I am doing the whole thing, it makes my brain work, my brain does not work as well as it did before I had a heart attack.
BMWs spark twice on each cylinder, it's called a wasted spark system...
Charles
2016 CCM 450
1975 R90/6
1958 R50
2012 R1200R
Image

User avatar
Rob Frankhamr
Club Member 13
Posts: 2973
Joined: Mon Jul 25, 2005 6:33 pm
Country of Residence: Scotland
Location: Kinloch Rannoch, Perthshire

Re: Electronic ignition

Postby Rob Frankhamr » Tue Jan 15, 2019 9:37 pm

I concur.

I keep thinking about whether or not I have the revs doubled but I don't think I have. Same applies to the actual angle, do I have it doubled or halved.

Thinking about this is one of the reasons I am doing the whole thing, it makes my brain work, my brain does not work as well as it did before I had a heart attack.
BMWs spark twice on each cylinder, it's called a wasted spark system...

Exactly, which is the same as saying once per revolution of the crankshaft... or the frequency of signal generated by Mjolinors test rig is at parity with the frequency of the ignition pulse for a given engine speed without dividing or multiplying by two...

Rob
Robin Frankham
ImageImageImage

Frankhams retirement home for elderly Boxers.

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Tue Jan 15, 2019 10:33 pm

I concur.

I keep thinking about whether or not I have the revs doubled but I don't think I have. Same applies to the actual angle, do I have it doubled or halved.

Thinking about this is one of the reasons I am doing the whole thing, it makes my brain work, my brain does not work as well as it did before I had a heart attack.
BMWs spark twice on each cylinder, it's called a wasted spark system...

Exactly, which is the same as saying once per revolution of the crankshaft... or the frequency of signal generated by Mjolinors test rig is at parity with the frequency of the ignition pulse for a given engine speed without dividing or multiplying by two...

Rob
But it has the Boyer shape when divided by two so I ballsed up somewhere but I can't see it.

I am more conducive to agreeing with the Boyer graph that is on the net now and am proceeding with it as a target.

Deep thinking going on right now about tying the specifics down. I have decided:

use one hall effect sensor
set it to detect at 49 degrees before TDC
undetect at 9 degrees before TDC

That gives me 19 degrees (400 us @ 7k) to crunch the numbers, plenty of time. It also allows me to set static timing as I put an LED on the Hall sensor PCB so when the LED is just going out and the 9 degree mark on the flywheel is lined up then the timing is correct.
It uses the detect off rising edge to fire the spark up to 1000 RPM then it will follow the programmed advance curve either with an equation or with a look up table. Both work OK and I haven't decided which is preferred.

Struggling at the moment with angular velocity variation with piston position, not sure at all how relevant it is and with number of entries I need in the look up table.

I have 64 bytes of storage. That gives me 60 values between 0 and 255 for the advance. I think 60 steps will be enough for the full advance curve and maybe half that would do.

Roy Gavin
Forum User
Posts: 410
Joined: Sun Jan 21, 2018 9:29 am
Country of Residence: Australia

Re: Electronic ignition

Postby Roy Gavin » Tue Jan 15, 2019 10:46 pm

Chas, looking at the Boyer graph it seems to me that that this unit has 32 degrees advance at 3200 rpm and 36 at 3500, and this looks like what Boyer is trying to achieve, not 36 at 32 which is what it would have if we follow the your instructions.
And that is single plug timing, not dual.
You don't say if the graph is for the analogue or Microdigital unit, I understand that the later units have a better/ different curve.
FWIW those who have checked on a dyno (RM) report that the optimum timing on a single plug system varies from bike to bike and is usually between 30 and 36 degrees at 3200 rpm, So if performance is important the stock figure is only to get you to the dyno!

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Thu Jan 17, 2019 9:43 pm

I am fixing the timing at 9 degrees if revs are below 1000.

Would it be sensible to retard to TDC if revs drop below 500 or so?

Roy Gavin
Forum User
Posts: 410
Joined: Sun Jan 21, 2018 9:29 am
Country of Residence: Australia

Re: Electronic ignition

Postby Roy Gavin » Fri Jan 18, 2019 3:13 am

Don't think so, easier just to set the butterfly stops set so the revs never drop below 1000, around where BMW recommend that they should be ---------.

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Fri Jan 18, 2019 9:37 am

I should have made it clearer. The purpose is to catch a stall and to improve cold start, not because your tick-over may be set incorrectly.

barryh
Forum User
Posts: 1045
Joined: Wed Jan 14, 2009 10:13 pm
Location: Cheshire, England

Re: Electronic ignition

Postby barryh » Fri Jan 18, 2019 10:24 am

I've always thought the kick in the boyer advance curve below 900RPM is a form of idle stabilisation that relies on the fact the normal idle timing is a compromise between easy starting and optimum timing for idle. i.e. it would idle faster with more advance. So if the revs fall the increasing advance below 900 RPM will tend to raise the idle speed back up again.

Mjolinor
Forum User
Posts: 1217
Joined: Tue Jun 12, 2018 9:38 pm
Country of Residence: United Kingdom
Location: Burnley

Re: Electronic ignition

Postby Mjolinor » Fri Jan 18, 2019 11:25 am

I don't grok that really.

As the engine slows there is a tendency to burn too much before TDC so you get a backfire. That is my thinking in retarding the ignition to TDC. At TDC it is not possible to backfire. 9 degrees before is a stand off between not enough force to shift bob weights and a good tick over. Ideally very little advance would be ideal but if you do that with a traditional system you cannot get a good advance curve. Electronic does not have that limitation.

Increasing advance at such low revs increases the chance of backfire and stall. It also means there is a higher minimum revs you must achieve before the engine will run than would be needed if the spark was at TDC.


Return to “Airhead Q&A's”

Who is online

Users browsing this forum: No registered users and 16 guests